Each year we at Carat write a trends presentation, looking at the themes we expect to become more prevalent and mainstream in the year to come.  You can find this year's presentation here, and listen to a webinar on the key trends for 2017 on 11th January.

It's also an opportunity to look back at what we predicted the year before, and how those themes have developed.

For example, one trend we saw was Challenges to Advertising, both in the rise of ad-free content like Netflix, and the rise of ad blocking.

In 2016 Netflix continued to grow, finishing the year at over 85m subscribers, and now available in all major counties, excluding China.  Amazon also pushed ahead with ad-free TV with the launch of a number of shows, including The Grand Tour, with Clarkson and friends, which broke Amazon's records for both streaming and subscription growth.  It now feels like a mainstream activity to watch this content, and marketers need to get smarter about how they reach audiences.

Ad blocking also rose, but it didn't reach the apocalypse that some predicted.  Yes, it's more popular than ever, but many publishers are now fighting back, refusing to allow people to see content if they have an ad blocker enabled, and in many cases this is making people either white-list certain sites, or pay for subscriptions.  Facebook, which isn't greatly affected by ad blocking as ads can't be blocked in its app, also started to develop anti ad-blocking tools, to make sure that desktop users would still see ads.

Another trend we looked at was Recognition – using the camera on smartphones for more than just taking pictures. 

Again, this rose, but often in ways that people didn't realise.  Face swapping, on apps like Facebook and Snapchat became very popular, and this is based on being able to identify objects and manipulate them.  Recognition is also one of the most important technologies in self-driving cars, and there are lots of applications in development around healthcare – identifying X-rays and more.

It's also fascinating that Snapchat has applied for a patent for targeting ads based in objects that it recognises in your pictures.  Again, we think that the possibilities are enormous for this trend.

Our trends last year – and what happened over the course of the year – were as follows…

Walled gardens

The big players have got even more powerful in 2016, with Google and Facebook taking a larger share of ad spend than before. They are also investing in content to strengthen their ecosystems - for example Twitter paying to stream NFL games.

Challenges to ads

Both ad-free content and ad blocking grew in 2016. Netflix subscribers grew to over 86m, while Amazon grew too and broke its own streaming records with 'The Grand Tour'. All of this content is ad-free. Ad blocking rose, but publishers fought back more effectively, taking a strong line against people who access sites with blockers.

The evolution of search

Amazon's Echo was one of the gadgets of the year, and has produced over 100 video ads showing off the voice search. Google introduced in-app search, Apple added paid search to the app store, and Snapchat bought Vurb, a search technology company.

Messaging & notifications

Both Whatsapp and Messenger hit 1bn active users in 2016, with WeChat coming in third globally at 700m users. Facebook enabled chatbots and games within its own Messenger, and more than 30,000 bots have been made to date.

Algorithms vs Curation

Algorithms, or gaming algorithms was one of the stories of the year, shown by how quickly fake news could spread on Facebook. Curation also continued to grow, partly driven by problems with algorithms - for example Facebook is now testing a curated news section, similar to Snapchat's Discover, which is only for approved content partners.

Maps & location

Maps were big news. As a location-based game Pokemon GO used maps heavily, bringing virtual creatures to real streets. TripAdvisor got into mapping by buying social mapping service CityMaps, and Google made a VR version of Google Earth for the HTC Vive.

Recognition

Facebook bought MSQRD, the popular face-swapping app. A Russian app, FindFace, went viral - it claims to be able to recognise people with 70% accuracy. Recognition is also being used in self-driving cars, and there is even a Messenger bot called 'WTF is that?' which will tell you what is in photos you send to it.

Sentiment

Facebook didn't introduce a way of targeting ads based on sentiment, but Twitter did. Snickers created a Hungerithm campaign, which used social media monitoring to change the price of the bars in Australia, dependent on national mood.

Dynamic Pricing

Price was an issue this year, but less than we expected. Uber still uses surge pricing (an urban myth says they use it when you have low battery and no alternative but to book), and Lidl experimented with dropping prices in the run up to Christmas in the UK, based on Tweet volumes.

'Telepathy'

We didn't see telepathy as such this year, but there were some interesting examples. For example, The Washington Post app can tell if you're bored and suggest different stories.